An overview of the economic impact of Indigenous disadvantage

Transcription

1 August 2008 An overview of the economic impact of Indigenous disadvantage Report by Access Economics Pty Limited for Reconciliation Australia

2 TABLE OF CONTENTS Glossary and Acronyms... iii Executive Summary... iv 1. Introduction Methodological Overview Gaps in Indigenous Data Report Structure Estimating GDP and Living Standards Introduction and Overview Parameters used to Estimate Lost Output What is being Modelled Summary of the Base Case and the what if scenario Modelling Approach Base Case Defined in Detail The what if Scenario Defined in Detail Calculated Model Inputs Results A Change in the Structure of Government Budgets Introduction Revenue Expenditure Burden of Disease Overview and Definitions Burden of Disease Excess Burden of Disease Among Indigenous Australians Conclusions References Appendix A: Calculation Method for Projections of Government Budget Scenarios Appendix B: Differences in estimates of the Life Expectancy Gap between Indigenous people and all Australians CHARTS Chart 2-1: Indigenous and non-indigenous population distributions, 2007 and 2026 the base case 8 Chart 2-2: Forecast life expectancy at birth by gender and population group base case 10 Chart 2-3: Comparison of range of industry specific average weekly earnings 14 Chart 2-4: Indigenous life expectancy and mortality distributions, 2007and Chart 2-5: What if scenario Indigenous population forecast and change in age distribution 16 i

3 Chart 2-6: What if scenario on employed Indigenous participation by industry 17 Chart 2-7: Change in total weekly industry earnings and average weekly earnings 18 Chart 2-8: Comparative populations income quintile* distribution by highest educational attainment level 21 Chart 2-9: Average weekly income based on highest level of educational attainment and equivalent income profiles 21 Chart 3-1: Total expenditure (all Australians) incorporated in this analysis, $ million, TABLES Table 2-1: Summary base case assumptions 6 Table 2-2: Summary of what was modelled (the what if scenario) 6 Table 2-3: Labour force participation, age standardised (% population 15 to 64) 10 Table 2-4: Relative income ratios for employed persons aged 15 years and over, median gross weekly individual income (a) 13 Table 2-5: Industry specific weightings by type of occupation 13 Table 2-6: Indigenous and Australian populations compared (number of people) 14 Table 2-7: Indigenous and Australian labour force compared (number of people) 17 Table 2-8: Inputs for AE-GEM model 18 Table 2-9: Results of AE-GEM analysis 19 Table 3-1: Impact on government revenue ($2009 million) 24 Table 3-2: Welfare projections and savings, (2009 dollars $ 000) (a) 28 Table 3-3: Health expenditure per capita, by broad health service and Indigenous status, and Table 3-4: Impact on government health expenditure in 2029(a) 31 Table 3-5: Attainment of national benchmarks, Table 3-6: Impact of education participation on government budgets(a) 33 Table 3-7: Impact of education related welfare on government budgets(a) 34 Table 3-8: Impact of justice on government budgets (a) 35 Table 3-9: Public and community housing related changes to government budgets (a) 37 Table 3-10: Rental assistance changes to government budgets (a) 38 Table 4-1 Rank of leading causes of DALYs, Indigenous Australian and total Australian populations, Table 4-2 Leading causes of premature death (YLLs) of Indigenous Australians, Table 4-3 Leading causes of years lived with disability of Indigenous Australians, Table 6-1 Differences in the ABS and AE life expectancy gap estimates 59 ii

4 GLOSSARY AND ACRONYMS ABS Australian Bureau of Statistics AE-DEM AE-GEM AIHW AWE CDEP CVD DALY DEEWR FaHCSIA GDP GST SCRGSP TWE VET YLD YLL Access Economics Demographic Model Access Economics General Equilibrium Model Australian Institute of Health and Welfare Average weekly earnings Community Development Employment Program Cardiovascular Disease Disability-Adjusted Life Years Department of Education, Employment and Workplace Relations Department of Families, Housing, Community Services and Indigenous Affairs Gross Domestic Product Goods and Services Tax Steering Committee for the Review of Government Service Provision Total weekly earnings Vocational Education and Training Years Lived with a Disability Years of Life Lost iii

5 EXECUTIVE SUMMARY Reconciliation Australia commissioned Access Economics to assess the economic benefits of alleviating Indigenous disadvantage including: 1 the impact on Gross Domestic Product (GDP) and living standards in Australia; and 2 the impact on the structure of government budgets. The aim is to provide a what if style scenario analysis to depict a hypothetical Australia where Indigenous Australians face the same opportunities and experiences as all Australians. The focus is on what could be. The complex policy and program mechanics of how to achieve a real reduction in the disadvantages faced by Indigenous Australians are not addressed. IMPACT ON GDP The analysis in this report shows there are sizeable economy wide benefits to be achieved from improving the quality of life of Indigenous Australians. In a what if scenario based on raising the life expectancy of Indigenous Australians and increasing the proportion of the Indigenous population in the workforce and who are able to take on higher skilled and better paid jobs to levels commensurate with those of all Australians, real GDP could be 1% higher than otherwise in 2029 equivalent to around $10 billion today. Further, since the increase in GDP is larger than the forecast increase in the total population, national living standards for all Australians would increase. There are therefore clear economic benefits from government action to reduce Indigenous disadvantage. The economic benefits will only be realised if the health and educational attainment of Indigenous Australians improves. In fact the modelling outcomes are predicated on the many facets of Indigenous disadvantage that contribute to their poorer health and labour market outcomes being addressed. In another light, achieving the economic benefits implies an improved quality of life for Indigenous people a reduction in the burden of disease and an improvement in the ability of Indigenous Australians to share in economic prosperity. These are generally not counted in economic indicators of progress but are nevertheless additional and no less important payoffs associated with the what if scenario modelled here. IMPACT ON THE STRUCTURE OF GOVERNMENT BUDGETS Estimates of the impact of improvements in Indigenous life expectancy, labour force participation and labour productivity on government budgets are summarised in the table below. These estimates do not include the costs of government programs designed to address problems of Indigenous disadvantage but rather illustrate the scale of the strengthening in government budgets that would flow from improved conditions for Indigenous Australians no matter the source of that improvement. If the circumstances of Indigenous Australians improve to match those of the Australian average: government revenue in 2029 would be $4.6 billion higher than otherwise; and government expenditure in 2029 in key portfolios relevant to Indigenous Australians would be $3.7 billion lower than otherwise. iv

6 SUMMARY OF GOVERNMENT BUDGET OUTCOMES (2009 $M) Government revenue (2009 $m) Income Income Tax $105,498 $124,786 $126,476 $1, % GST $49,927 $63,816 $64,350 $ % Corporate tax $67,467 $79,801 $80,577 $ % Other Revenues $187,726 $234,514 $236,579 $1, % TOTAL REVENUE $410,618 $502,917 $507,981 $4, % Government Budgets (2009 $m) Baseline revenue 2009 Baseline revenue 2029 Baseline Indigenous Expenditure Life Expectancy Shock revenue 2029 Indigenous Shock Expenditure (2029) LE & Labour Force* Increase in revenue 2029 Combined shocks Overall Savings Increase in 2029 due to shock Increase in 2029 due to shock Welfare Newstart Allowance $404 $629 $660 $156 $473 75% Youth Allowance $93 $111 $114 $21 $90 81% Parenting Payment Partnered $84 $127 $133 $39 $88 69% Parenting Payment Single $330 $496 $519 $168 $328 66% Sickness Allowance $2 $3 $4 $3 $1 21% Disability Support $376 $550 $575 $321 $230 42% Aged Pension $129 $334 $429 $563 -$229-68% Carer's Payment $65 $98 $104 $56 $42 43% Carer's Allowance $60 $91 $97 $52 $39 43% CDEP $120 $175 $185 $5 $170 97% Total $1,663 $2,613 $2,818 $1,383 $1,231 47% Health Admitted patient services - private hosp $24 $35 $37 $132 -$98-280% Admitted patient services - public hospi $1,425 $2,077 $2,193 $1,071 $1,007 48% Medical sevices $194 $283 $298 $602 -$ % Community health services $686 $1,000 $1,056 $183 $817 82% Dental services and other professional $49 $71 $75 $83 -$12-17% Pharmaceuticals $100 $145 $153 $314 -$ % Services for older people $41 $60 $64 $228 -$ % Patient transport $141 $205 $216 $73 $132 65% Public health activities and research $169 $247 $260 $144 $102 42% Other health services $106 $154 $163 $109 $45 29% Total $2,935 $4,278 $4,515 $2,939 $1,339 31% Education Preschool $101 $150 $154 $253 -$103-68% Primary and Secondary $1,411 $1,710 $1,767 $1,825 -$115-7% Tertiary (Higher Education) $196 $309 $330 $527 -$219-71% Vocational Employment Training (VET) $212 $333 $356 $145 $188 57% IESP $52 $73 $77 $77 -$4-5% IEDA $68 $106 $112 $112 -$6-6% Abstudy $208 $319 $341 $80 $239 75% Austudy and Youth Allowance $13 $20 $22 $7 $13 66% Total $2,262 $3,021 $3,160 $3,026 -$5 0% Justice Open plus periodic detention $136 $205 $219 $19 $186 91% Secure $430 $650 $694 $59 $592 91% Community Corrections $51 $77 $82 $8 $69 89% Total $616 $933 $996 $87 $846 91% Housing Indigenous community housing $63 $92 $97 $97 -$5-6% National community housing $17 $24 $25 $9 $15 63% State owned and managed indigenous $85 $124 $131 $131 -$7-6% National public (State owned) housing $160 $233 $246 $63 $170 73% Commonwealth Rental Assistance $76 $115 $123 $76 $39 34% Total $401 $588 $622 $375 $213 36% TOTAL EXPENDITURE $7,876 $11,433 $12,111 $7,810 $3,623 32% TOTAL AVAILABLE TO GOVERNMENTS EACH YEAR FROM 2029 $8,250 LFP = labour force participation rates * LE and LFP combined represents the move by the indigenous population to average expenditure per capita as the Indigenous LFP approaches the Australian average. That is, as indigenous labour force status improves the requirement for welfare decreases. Similarly, as education participation increases, labour force status also improves. v

7 Conclusions Foreshadowing possible policies and programs required to achieve the economic benefits was out of scope for this project. However, the analysis of government budgets suggests that from 2029, there will be an additional $8.3 billion available to governments each year if Indigenous disadvantage were alleviated. In principle, these additional public funds could be allocated to policies and programs aimed at improving the quality of life of Indigenous Australians. Monitoring and evaluation are necessary accompaniments of such an approach. If potential future benefits are used to justify upfront investments in education and health infrastructure, interim targets for the health and education workforce and complementary infrastructure, in concert with interim objectives measured in terms of outcomes for Indigenous people for example, child health and educational achievement are vital in tracking whether the benefits are being achieved. The Australian Government s current approach includes interim objectives to halve the literacy and numeracy achievement gap within a decade, halve the gap in employment outcomes within a decade, and provide Indigenous children with access to quality preschool programs within 5 years (Macklin, 2008). To some extent, these interim objectives are augmented with provisional infrastructure targets, although the latter could be strengthened with a view to increasing certainty that the potential economic benefits and improvements to Indigenous people s wellbeing will be achieved in the foreseeable future. Access Economics August 2008 vi

8 1. INTRODUCTION In February 2008, Reconciliation Australia commissioned Access Economics to assess: 1 the impact on GDP and living standards in Australia if the circumstances of Indigenous people were improved and reflected that of the all-australian average in terms of life expectancy, labour force participation and earnings; 2 how this might affect the structure of government budgets; and 3 how the health related quality of life of Indigenous people might change. A large number of reports have amassed evidence of the poor living standards, quality of life and socioeconomic outcomes of Indigenous people relative to other Australians. 1 While the costs of Indigenous disadvantage are sustained by Indigenous individuals and communities, there are also broader costs to the Australian society and economy. This report focuses on the tangible economic costs the benefits that would otherwise accrue if Indigenous Australians lived as long as all Australians, were able to participate in the workforce to the same extent and had the same earning capacity. These tangible economic costs can be measured in terms of foregone output and economic wellbeing (output per Australian). The aim of this project is to provide a what if style scenario analysis to depict a hypothetical Australia where Indigenous Australians face the same opportunities and experiences as all Australians. The focus is on what could be. The potentially complex policy and program mechanics of how to achieve a real reduction in the disadvantages faced by Indigenous Australians are not addressed here. 1.1 METHODOLOGICAL OVERVIEW The project was split into three parts. 1 Estimating the impact on GDP and living standards. The Access Economics General Equilibrium Model (AE-GEM) was used to forecast the impact on GDP and living standards in Australia if Indigenous life expectancy, Indigenous labour force participation rates and Indigenous productivity rates increased over 20 years so that they matched the national Australian average in This was compared with a base case in which Indigenous life spans, labour force participation rates and labour productivity remained at current levels for the next 20 years. 2 A change in the structure of government budgets (revenue and expenditure). The AE-GEM modelling outputs were used to estimate the impact of improvements in Indigenous life expectancy, labour force participation, and labour productivity on government revenue (Australian, State and Territory). The impact on government expenditure on mainstream programs was calculated by estimating current expenditure per head (Indigenous and all Australians) and adjusting the relevant population of Indigenous people for changes in life expectancy, participation in education or other relevant variables. Welfare payments (including employment payments such as Newstart Allowance, and other payments such as the Disability Support Pension and Aged Pension), health system expenditure, education expenditure, housing (including Commonwealth Rent Assistance) and justice system expenditure were included in the analysis. 1 For example, SCRGSP 2007 and ABS/AIHW

9 3 Burden of disease: The University of Technology, Queensland, produced a report on the burden of disease among Indigenous people (Vos et al, 2007). Their estimates of the excess burden of disease experienced by Indigenous Australians are summarised here. 1.2 GAPS IN INDIGENOUS DATA As part of this project, Reconciliation Australia requested the identification of data gaps for Indigenous Australians. A number of gaps have been itemised in other reports (see for example, various ABS information papers, SCRGSP 2007 and ABS/AIHW 2005). Some gaps are being rectified gradually over time (for instance, inclusion of Indigenous status in Medicare data collections). For this project, additional information that facilitated comparisons between Indigenous and non-indigenous people over time about the following would have been useful: hours worked by occupation and industry; employment income by occupation and industry; Indigenous migration patterns, between States and Territories and internationally; government revenue and the contribution of Indigenous Australians; consistency in reporting educational participation rates (age groups and time series); persons per household for community housing; and epidemiological information. Vos et al, (2007) noted that available data for Indigenous Australians is limited and data sources allowing direct measurement of disease parameters for Indigenous Australians are few. Those which are available are often not representative for all Indigenous Australians. 1.3 REPORT STRUCTURE The report reflects the three components of the project. Chapter 2 details the approach and findings of the general equilibrium modelling. Potential changes in the structure of government budgets are discussed in Chapter 3. Health related quality of life and the excess burden of disease experienced by Indigenous Australians are outlined in Chapter 4. Brief conclusions about the implications of these three components are made in Chapter 5. Note: consistent with the approach of the Australian Bureau of Statistics and others, the term Indigenous in this report encompasses Aboriginal and Torres Strait Islander Australians. 2

10 2. ESTIMATING GDP AND LIVING STANDARDS The discussion in this chapter focuses on the methodology and findings of the general equilibrium modelling used to estimate the additional output that would accrue if the Indigenous life span, Indigenous labour force participation rates and earnings rates were the same as the Australia-wide average by Projections are generated for 20 years and beyond (2009 to 2029+) reflecting views about the length of a generation. 2.1 INTRODUCTION AND OVERVIEW First, it is important to note that the discussion in this Chapter includes references to the years 2026 as well as The reasons are outlined in Box 2-1. Box 2-1 Explanation of why this Chapter refers to both 2026 and 2029 Throughout this chapter references are made to the years 2026 and While this can cause confusion, it reflects the nature of the parameters used as inputs to the modelling, and our modelling aims. The modelling objective was to project gross domestic product (GDP) 20 to 30 years into the future. This projection period was chosen based on an approximation of the length of a generation. However, projections of the Indigenous population were only available to 2026 (because of the underlying fertility, mortality and other input projections available). The estimates made in this paper therefore assume the gap in life expectancy between Indigenous people and all Australians is closed by the year Nevertheless, the modelling generates projections to 2029 and The analysis of government budgets in Chapter 3 is also based on generating positive change within a generation (2009 to 2029) and is consistent with the modelling in this Chapter. Previous international and Australian studies have estimated the output foregone as a result of Indigenous people having less education and poorer health outcomes and thus lower employment and productivity rates. Taylor and Stanley (2005) calculated that, in the Thamarrurr region of the Northern Territory (NT), the opportunity cost of the status quo for Indigenous people (the output gap) was $43.8 million per annum in In Thamarrurr, 82% of incomes were derived from welfare payments, and the median age at death was 46 years (Taylor and Stanley 2005:6-7). The output gap was calculated by applying NT average employment rates and average output per worker to the population of the Thamarrurr region. This methodology follows that of the Canadian Royal Commission on Aboriginal Peoples (1996), where the opportunity cost was measured in terms of national production foregone as a result of the continued poor socioeconomic status of Indigenous people. 2.2 PARAMETERS USED TO ESTIMATE LOST OUTPUT The Indigenous population, labour force participation and productivity were selected as parameters for economic modelling. Other things constant, if more Indigenous people 3

11 participate in the workforce 2, if Indigenous people are able to expand the number of hours they work per year, or if they produce more output per person per year, the higher the overall potential output produced and the higher Australian living standards (GDP per head). Life expectancy was used as a proxy for the Indigenous population, and average weekly earnings as a proxy for labour productivity. Improvements in life expectancy, workforce participation and productivity depend on addressing the many facets of Indigenous disadvantage broadly defined as health and educational outcomes. Health and education are not explicitly modelled to avoid double counting, but are nevertheless the key to improving the three modelled parameters (see for example, Box 2-2). 2 People who are employed or actively looking for work are participating in the labour force. More Indigenous people in the workforce may be achieved by an increase in the Indigenous population or an increased propensity to participate. 4

12 Box 2-2 Indigenous health and workforce participation Illness and disease can substantially reduce workforce participation and productivity. For example, the Productivity Commission (2007) cited studies that found, compared with the all-australian average: people with mental illness had workforce participation rates close to 30 percentage points lower; serious injury reduced participation rates by 13 percentage points; type 2 diabetes reduced participation rates by around five percentage points; and CVD reduced participation by around three percentage. The top seven broad causes of the Indigenous disease burden in 2003 in order from the highest were CVD, mental disorders, chronic respiratory disease, diabetes, cancers, unintentional injury, and intentional injury. The largest differentials in DALYs per 1,000 people between Indigenous and all Australians were for CVD, diabetes mellitus and intentional injuries in both males and females (Vos et al 2007:30). Figure 4.4 from Vos et al (2007) is reproduced here. Preventing and ameliorating illness and disease among Indigenous Australians is likely to substantially improve Indigenous workforce participation. Source: Productivity Commission (2007) and Vos et al (2007). 2.3 WHAT IS BEING MODELLED SUMMARY OF THE BASE CASE AND THE WHAT IF SCENARIO Lost output is estimated by comparing the base case with the what if scenario. A summary of the base case and what if scenario is provided here. The base case is outlined in more detail below in Section 2.5 and the what if scenario is outlined in more detail in Section 2.6. The base case assumptions are summarised in Table

13 TABLE 2-1: SUMMARY BASE CASE ASSUMPTIONS Population Labour force participation Australian average Life expectancy increases based on historical average for last 100 years Constant at current rates Indigenous people No change in mortality rates so life expectancy constant at current levels Constant at current rates Labour productivity AWE= average weekly earnings. Industry specific AWE increases based on historical average wage inflation rate for the last 10 years Gap between Indigenous and Australian average income, adjusted for full-time/part-time differences and, increasing at 0.4% per annum based on increase in historical disparity in income over the 5 years to 2001, applied to Australian average AWE. In the what if scenario, the life span of Indigenous Australians is gradually increased over time so that it matches the Australian average by Indigenous labour force participation and labour productivity also gradually increase to match the Australian average by The modelled what if scenario is summarised in Table 2-2. TABLE 2-2: SUMMARY OF WHAT WAS MODELLED (THE WHAT IF SCENARIO) Benchmark Average Australian life expectancy in 2026 Average Australian labour force participation rates in 2026 Average Australian labour productivity in 2026 AWE= average weekly earnings. Change Indigenous life expectancy matches the benchmark by 2026 Indigenous labour force participation rates match the benchmark by 2026 Indigenous labour productivity matches the benchmark by 2026 Indicator Australian average annual population growth rate Australian wide average industry specific labour force participation rates Australian average industry specific AWE (increasing at rate of annual wage inflation) 2.4 MODELLING APPROACH The modelling approach is based on a combination of two forecasting constructs. The first is an excel-based model used to predict the impact of changes in the prescribed parameters on a range of macroeconomic indicators (relating to the entire population) over a 20 year time horizon (refer to Table 2-2 for details). The impact on each macroeconomic indicator is expressed as the difference between the base case (or status quo ) and the what if scenario annualised forecast results. The Indigenous population what if forecast is based on a basic linear model (or constant annual growth rate) over the 20 year time horizon. The resultant annual changes in each of the three macroeconomic indicators are then used as inputs into AE-GEM (Access Economics - General Equilibrium Model). AE-GEM is a large scale, dynamic, multi-region, multi-commodity computable general equilibrium model of the world economy, with an explicit representation of the Australian economy. The model projects changes in macroeconomic aggregates, such as GDP, employment, export 6

14 volumes, investment and private consumption, and sectoral aggregates, such as output, exports, imports and employment. Using AE-GEM, we estimate the deviation from a business-as-usual scenario of the changes in the three macroeconomic indicators listed in Table BASE CASE DEFINED IN DETAIL POPULATION Access Economics collects historical data from ABS publications which it uses to forecast both total Australian and Indigenous Australian population by age. The Access Economics demographic model (AE-DEM) is updated quarterly utilising a dynamic modelling approach, and based on historical national level data, relating to: fertility rates; life expectancy from birth; age specific mortality distributions; and international migration rates. The Access Economics Indigenous demographic model (AE-DEM-I) incorporates the same dynamic methodology, and is based on the fertility, mortality rates, and life expectancy of Indigenous Australians. Since there is a paucity of data relating to Indigenous international migration rates, the model has assumed zero Indigenous international migration. While this may appear to be an underestimation of the actual trends, we do not expect this assumption to be influential in the overall results. Chart 2-1 presents gender specific population age distributions for the base case as at 30 June 2007 and 30 June The Indigenous population declines markedly with age, despite Indigenous historic and forecast fertility rates exceeding those of non-indigenous people. The difference reflects higher mortality rates and lower life expectancy of Indigenous people than other Australians. Overall, at 30 June 2007, Indigenous Australians accounted for 2.5% of the Australian population (2.5% of Australian males and 2.5% of Australian females). By 30 June 2026, in the base case, Indigenous people account for 3.0% of the Australian population (2.7% of Australian males and 3.0% of Australian females). The greater discrepancy between the relative proportions of Indigenous males and females reflects gender differences in the life expectancy of Indigenous people. 7

15 CHART 2-1: INDIGENOUS AND NON-INDIGENOUS POPULATION DISTRIBUTIONS, 2007 AND 2026 THE BASE CASE Females Indigenous 50,000 40,000 30,000 Non-indigenous 1,000,000 Indigenous 2007 Indigenous 2026 Non-indigenous ,000 Non-indigenous ,000 20, ,000 10, , year age groups 0 Males Indigenous 50,000 40,000 30,000 Non-indigenous 1,000,000 Indigenous 2007 Indigenous 2026 Non-indigenous ,000 Non-indigenous ,000 20, ,000 10, , year age groups Source: AE-DEM In the base case, the comparatively high fertility rate and low life expectancy of Indigenous Australians results in a relatively young and dependent Indigenous population cohort. At 30 June 2007, the median age of the national population was 35.8 years (38.9 by 30 June 2026) compared with only 20.2 years (24.2 by 30 June 2026) for Indigenous people. At 30 June 2007, the dependency ratio (the population aged less than 16 years plus those over 64 years as a proportion of the working age population 16 to 64 years) was around 51% for all Australians compared to around 77% for Indigenous Australians. 8

16 2.5.2 LIFE EXPECTANCY AT BIRTH Historical life expectancies for both the Indigenous population (ABS 2004a) and total Australian population (ABS 2004a) were obtained from the ABS. In the base case, there is no change in the mortality rates for the Indigenous population during the entire projection period, consistent with ABS (2004a). This means that the estimate of Indigenous life expectancy at birth is projected to be constant at the level (ABS 2004). For the total population, the ABS has data going back 100 years relating to life expectancy at birth. Over the past century, male and female life expectancy at birth has increased by 23.5 years and 24.7 years respectively. These trends are allowed for in the forecast model, in line with ABS medium population projections (i.e. life expectancy increases over time, but the rate of increase falls over time). Box 2-3 Justification for the base case assumption of no change in Indigenous life expectancy over the 20 year projection period An individual s status as an Aboriginal or Torres Strait Islander is not always identified when deaths are registered, resulting in inaccurate assessments of deaths and mortality rates of Indigenous Australians. The ABS estimated that Indigenous age-specific death rates may understate the true rates by as much as 55%. Fluctuations in historical Indigenous mortality rates can be partly attributed to the changing levels of coverage of Indigenous deaths. Given this volatility, historical Indigenous mortality trends should be viewed with caution. The ABS (2006d) conducted sensitivity analysis on their projections of the Indigenous population for the period 2002 and 2009 under different mortality assumptions, all else equal. The base case was no change in mortality rates (constant over time). The ABS (2006d) concluded that a reduction in mortality rates that resulted in an annual improvement in Indigenous life expectancy at birth of 0.25 years would have only a minimal impact on the projected Indigenous population in Since there are no reliable data with which to inform Indigenous life expectancy growth forecasts, and consistent with the ABS (2006d), the assumption of no change in life expectancy at birth is employed for this analysis. ABS (2004a) estimated the life expectancy at birth for people of Indigenous origin born in the period was 59.4 years for males and 64.8 years for females. The ABS (2004a) compared this with the life expectancy for all Australian males and females of 76.2 years and 81.8 years respectively for persons born in the period and 77.4 years and 82.6 years respectively for persons born in the period The ABS (2004) concluded that, This indicates that the life expectancy of Indigenous population is approximately 18 years less than that of the total Australian population (ABS 2004a:15). For this project, while as previously discussed, the life expectancy at birth of Indigenous Australians is assumed constant and based on the life expectancy at birth for people of Indigenous origin born in the period , the life expectancy at birth for all Australians is updated to more recent ABS estimates. For all Australian males and females, the ABS (2006b) estimated that life expectancy at birth for those born in the period was 78.7 years and 83.5 years respectively. Based on these updated data, in 2007, the gap in 9

17 life expectancy between Indigenous people and all Australians is approximately 19 years (Chart 2-2). The reasons for differences in estimates of the life expectancy gap between Indigenous Australians and all Australians are discussed in more detail in Appendix B. CHART 2-2: FORECAST LIFE EXPECTANCY AT BIRTH BY GENDER AND POPULATION GROUP BASE CASE Female LE Gap = 18.7 yrs Male LE Gap = 19.3 yrs Indig (Male) Total (Male) Indig (Female) Total (Female) Female LE Gap = 22.6 yrs Male LE Gap = 24.1 yrs Forecast Year (as at 30 June) Source: AE-DEM LABOUR FORCE PARTICIPATION Labour force participation is defined as people working or willing to work aged 15 to 64 years old. AE-GEM provides for a split by industry. Estimates of Indigenous labour force participation suggest that in the three years to 2005, Indigenous participation grew 3.6%, whereas the national average grew 5.7%. However, in the 10 years to 2005, Indigenous labour force participation decreased by 1.0%. Implying, as expected, that labour force participation is closely linked to the positioning of economic and business cycles at the time. It is therefore reasonable to assume in the base case that, for both the Indigenous and national average, labour participation rates are constant over time. Estimates of labour force participation rates were obtained from SCRGSP (2007) (Table 2-3). In the base case, labour force participation and productivity estimates are modelled inclusive of CDEP participants, in line with ABS methodology (Box 2-4). TABLE 2-3: LABOUR FORCE PARTICIPATION, AGE STANDARDISED (% POPULATION 15 TO 64) Indigenous non-indigenous National average Male 67.8% 85.6% 85.2% Female 50.2% 70.7% 70.3% Total 58.5% 78.1% 77.7% Source: SCRGSP 2007, CDEP participants included. 10

18 Box 2-4 Community Development Employment Program (CDEP) The Community Development Employment Project (CDEP) is a program for unemployed Indigenous Australians, funded by the Australian Government. The aim is to develop the workplace skills of participants. The program was to be abolished in mid-2008, but the new Government placed a moratorium on dismantling it in December CDEP has already been wound up in urban and regional centres, where it had a relatively minor role. However, CDEP remains significant in remote and very remote areas. In , CDEP accounted for 68.0% of Indigenous employment in very remote areas (SCRGSP 2007:11.6). There were 34,775 CDEP participants as at 30 June 2005 (SCRGSP 2007:3.43). Some CDEP programs provide essential services such as health care and education (SCRGSP 2007 p.41). Categorisation of CDEP for the purposes of defining the labour force is therefore difficult. The ABS includes CDEP participants in estimates of the labour force and employment EMPLOYMENT To derive employment rates, the labour force participation rates were adjusted for unemployment using Access Economics Business Outlook (2008), which publishes forecasts of average Australian unemployment rates to 30 June The unemployment rate for the remainder of the projection period (2017 to 2026) was assumed constant at the 4.4% 2017 forecast rate. An Australian average unemployment rate was derived based on a weighted average of the Indigenous and non-indigenous rates (96%) using comparative data from SCRGSP (2007). The ratio of employed Indigenous labour force to employed national labour force is used to adjust national and Indigenous unemployment over time. Employment estimates were then adjusted for differences between the Indigenous and Australian full-time/part-time work split. Indigenous and non-indigenous employment rates by full-time versus part-time as at 30 June 2005 were based on SCRGSP (2007). The method is illustrated in Box

19 Box 2-5 Method for calculating full-time equivalent employment (FTE) estimates In 2005, 40% of employed Indigenous people worked part-time (SCRGSP 2007). These data are assumed constant throughout the forecast period. Projections for this report suggest 176,161 Indigenous people will be employed in 2009, of which 19,828 will be participating in CDEP. (CDEP participant numbers were not able to be adjusted for hours worked so are effectively treated as full-time equivalents.) FTE employment levels for the remainder of the work force were estimated on the basis that a part-time worker works half as many hours as a full-time worker. (This assumes that part-time hours per week are distributed normally.) The calculation is therefore: (176,161-19,828) x (0.5 x 40% x 60%) = 125,219 The estimate of total FTE Indigenous employment is 145,047, including CDEP participants. (Note that the calculation does not reconcile exactly due to rounding.) LABOUR PRODUCTIVITY As indicated in Section 2.3, the difference between the Average Weekly Earnings (AWE) of Indigenous and non-indigenous people adjusted for differences in hours worked is used as an indicator of the difference in labour productivity. ABS 2001 Census data (ABS 2003) were used to estimate the difference between Indigenous and Australian average AWE by occupation 3 (Table 2-4). The estimates were adjusted to reflect the proportions of full-time and part-time employment to derive a relative FTE income ratio by occupation (Table 2-4). Notably: Between 1996 and 2001, average equivalised gross household income for Indigenous persons rose by about 11% (after adjustment for inflation using the Consumer Price Index) compared with 13% for non-indigenous persons. (ABS 2003) Therefore, the model assumes a per annum increase in the relative (Indigenous versus national average) income gap of 0.4% (or 2 percentage points over 5 years). This factor is applied to the 2001 income ratios to achieve income ratios for The income measure used to identify the gap in income is gross individual income per week for employed people. While the measure is before tax and other deductions and may include welfare payments that top up the incomes of employed people, the scope is the same for both Indigenous and non-indigenous people. The ABS (2003) defined this measure as: the usual gross weekly income of persons aged 15 years or over. Gross weekly income is income before tax, superannuation, health insurance, or other deductions are made, and includes family payments, pensions, unemployment benefits, student allowances, maintenance (child support), superannuation, wages, overtime, dividends, rents received, interest received, business or farm income (less operating expenses) and workers compensation received. 12

20 TABLE 2-4: RELATIVE INCOME RATIOS FOR EMPLOYED PERSONS AGED 15 YEARS AND OVER, MEDIAN GROSS WEEKLY INDIVIDUAL INCOME (A) (A) See footnote 3. (B) Adjusted to FTE earnings based on employment type breakdown and a 0.5 weighting on part-time workers (refer to Section for explanation). Source: ABS Data is inclusive of CDEP participants. The AWE ratios for occupations are applied to each industry on a weighted average basis (using ABS 2001 Census estimates of employment by occupation in each industry) (Table 2-5) to obtain industry specific AWE estimates for Indigenous AWE estimates by industry are estimated as the product of the Australian AWE estimates by industry and the AWE Indigenous income ratios by industry. TABLE 2-5: INDUSTRY SPECIFIC WEIGHTINGS BY TYPE OF OCCUPATION Source: ABS 2001 Census employment by occupation by industry Two methods were used to obtain forecasts, as detailed below. National Average 2001 Indigenous 2001 (adj) (B) National Average 2001 (adj) (B) Indigenous 2001 Income Ratio 2001 Managers and administrators $740 $912 $1,035 $1,176 88% 86% Professionals $639 $876 $893 $1,130 79% 77% Associate professionals $606 $695 $847 $896 95% 93% Tradespersons and related workers $492 $596 $688 $769 89% 87% Advanced clerical and service workers $537 $582 $751 $ % 98% Intermediate clerical, sales and service workers $443 $503 $619 $649 95% 93% Intermediate production and transport workers $521 $567 $728 $ % 98% Elementary clerical, sales and service workers $302 $319 $422 $ % 101% Labourers and related workers $238 $420 $333 $542 61% 59% Not stated/inadequately described $247 $482 $345 $622 56% 54% Total $431 $587 $603 $757 80% 78% Personal & Other Services Construc Manuf Retail Income Ratio 2006 Property & Business Services Other Total Gov't & Defence Managers and administrators 10% 4% 10% 12% 4% 8% 11% 9% Professionals 22% 12% 3% 9% 3% 34% 24% 18% Associate professionals 14% 21% 7% 5% 14% 15% 12% 12% Tradespersons and related workers 8% 23% 48% 25% 13% 3% 5% 12% Advanced clerical and service workers 3% 3% 5% 3% 2% 9% 3% 4% Intermediate clerical, sales and service wo 26% 15% 4% 9% 10% 11% 22% 16% Intermediate production and transport work 4% 5% 9% 17% 7% 2% 8% 8% Elementary clerical, sales and service work 6% 9% 1% 2% 40% 6% 5% 10% Labourers and related workers 6% 8% 12% 16% 6% 10% 7% 9% Not stated/inadequately described 2% 1% 1% 1% 0% 1% 1% 1% Total 0% 0% 0% 0% 0% 0% 2% 1% Australian-average AWE historical trends by industry were sourced from the ABS (ABS 2008a). Forecast Australia average industry specific AWE were then based on the historical data being inflated by the average annual wage inflation rate over the last 10 years. The quarterly Access Economics Business Outlook report (2008) publishes forecast Australian average AWE over a 10 year time horizon (to 30 June 2017). These forecasts are reconciled against the industry specific Australian average AWE forecasts defined in the previous bullet point. The Other industries sector, a conglomerate of a number of ABS industry definitions, is used as the balancing item. 13

21 CHART 2-3: COMPARISON OF RANGE OF INDUSTRY SPECIFIC AVERAGE WEEKLY EARNINGS $2,500 $2,000 High (Construc.) National Average Low (Retail) High (Construc.) Indig Average Low (Retail.) National AWE Range $1,500 $1,000 $500 Indigenous AWE Range $ Source: AE Indigenous Supply Model 2.6 THE WHAT IF SCENARIO DEFINED IN DETAIL As summarised in Table 2-2, the modelling aims to assess the impact of increasing the life span, labour force participation rates and labour productivity of Indigenous people to match that of the Australian average by POPULATION To achieve longer life spans for Indigenous people, AE-DEM was used to generate projections of the Indigenous population by adjusting Indigenous male and female age specific mortality rates so that they reflected the Australian average. The increase in the life expectancy of Indigenous Australians means that by 2026, there are 36,750 more Indigenous people in the what if scenario compared with the base case. The average annual what if population growth rate (1.84%) is 0.2 percentage points higher than that of the base case (1.66%). Table 2-6 summarises the differences in populations between the base and what if scenario cases. To put these changes in the Australian population in context, AE-DEM forecasts of increases in the Australian population due to international immigration (based on historical estimates) for 2006 are 146,753 and for 2026 are 171,664. Year TABLE 2-6: INDIGENOUS AND AUSTRALIAN POPULATIONS COMPARED (NUMBER OF PEOPLE) Indigenous Base case What if scenario Difference Total Australian Indigenous Total Australian ,196 21,017, ,347 21,017, ,289 26,719, ,038 26,755,979 36,749 Source: AE-DEM 14

22 CHART 2-4: INDIGENOUS LIFE EXPECTANCY AND MORTALITY DISTRIBUTIONS, 2007AND 2026 Indigenous Mortality Distributions Mort. Dist. (F) 2007 Mort. Dist. (M) 2026 Mort. Dist. (F - shock) 2026 Mort. Dist. (M - shock) Birth Age Indigenous Life Expectancy Distributions Life Exp. (F) 2007 Life Exp. (M) 2026 Life Exp. (F - shock) 2026 Life Exp. (M - shock) Forecast Year Source: AE-DEM 15

23 CHART 2-5: WHAT IF SCENARIO INDIGENOUS POPULATION FORECAST AND CHANGE IN AGE DISTRIBUTION Indigenous Population Forecast 900 Annual Base Population Annual Shock Population Annual Population ('000) Forecast Year Indigenous Age Distributions Age Specific Population ('000) Age Dist Age Dist. (shock) 2026 Age Dist. (base) Year Age groups Shock series represents the what if scenario the impact of both the life expectancy and labour force participation rate LABOUR FORCE PARTICIPATION The increase in Indigenous labour force participation rates, combined with the rise in Indigenous life expectancy, means that by 2026, there are 104,072 more Indigenous people in the labour force in the what if scenario compared with the base case. The average 16

24 annual what if scenario population growth rate (1.22%) is 0.01 percentage points higher than that of the base case (1.23%). Table 2-7 summarises the differences in the number of people in the labour force between the two cases. TABLE 2-7: INDIGENOUS AND AUSTRALIAN LABOUR FORCE COMPARED (NUMBER OF PEOPLE) Year Base case What if scenario Difference Indigenous Total Australian Indigenous Total Australian ,086 11,028, ,614 11,031,422 2, ,303 13,051, ,375 13,155, ,072 Source: AE-DEM and calculations by Access Economics. Chart 2-6 shows the differences in the base case and what if scenarios. In particular, note the different distribution of employment by industry in In the model, the proportion of Indigenous people employed in the government sector and in the personal services industry is lower than in the base case, and the proportion employed in construction, manufacturing, retail trade and business services sectors is higher (reflecting the distribution for all Australians). CHART 2-6: WHAT IF SCENARIO ON EMPLOYED INDIGENOUS PARTICIPATION BY INDUSTRY 100% Participation Rate = 55% Participation Rate = 59% Participation Rate = 77% 80% 60% 40% 20% 0% Base Case Shock Base Case Shock Base Case Shock Base Case Shock Base Case Shock Govt Personal Services Construction Manufacturing Retail Business Services Other Other industries include mining, wholesale trade, utilities, transport and storage, and communication services. Shock series represents the what if scenario the impact of both the life expectancy and labour force participation rate LABOUR PRODUCTIVITY Given all targets are achieved by 2026, the estimated Total Weekly Earnings (TWE) for the Indigenous population in 2026 will have exceeded the base case by $26.8 million (in 2026 dollars). 17

25 CHART 2-7: CHANGE IN TOTAL WEEKLY INDUSTRY EARNINGS AND AVERAGE WEEKLY EARNINGS TWE ($m) Govt (LHS) Health Services (LHS) AWE ($) $600 Personal Services (LHS) Construction (LHS) $2,000 Manufacturing (LHS) Retail (LHS) Business Services (LHS) AWE (Base) $450 AWE (Shock) $1,500 History Forecast $300 $1,000 $150 $500 $0 Base Case Shock Base Case Shock Base Case Shock Base Case Shock Base Case Shock $0 The Shock series represents the what if scenario the impact of life expectancy, participation rates, and AWE shocks. Total Weekly Industry Earnings (TWE) is estimated by AWE and participation numbers for each industry. Mini sensitivity analysis found that As part of the model it is assumed that the distribution of Indigenous employment by industry matches that of the national average by Sensitivity analysis was undertaken to test the impact of leaving the distribution of Indigenous employment by industry the same in 2026 as it was in There was no appreciable impact on the results. 2.7 CALCULATED MODEL INPUTS Table 2-8 details the annual growth rates in the Australian population, the labour force and labour productivity in the base case, and those required to achieve the what if scenario benchmarks (see Table 2-2). The deltas are the differences in annual growth rates between the base case and what if scenario. These are the inputs for the AE-GEM modelling. Source: Calculations by Access Economics TABLE 2-8: INPUTS FOR AE-GEM MODEL Labour force, life expectancy and Income shock impact analysis Current Status Shock Status Delta Current Status Shock Status Delta Annual population growth rate 1.22% 1.23% 0.01% Industry Annualised Data Employed Labour supply growth p.a. Wage inflation growth on AWE p.a. Government, Defence, Health & education 0.9% 0.9% -0.02% 3.3% 3.3% 0.02% Personal and Other Services 0.9% 0.8% -0.13% 3.5% 3.5% 0.03% Construction 0.9% 1.0% 0.06% 4.1% 4.1% 0.01% Manufacturing 0.9% 1.0% 0.07% 3.4% 3.4% 0.01% Retail Trade 0.9% 1.0% 0.10% 3.8% 3.8% 0.00% Property and Business Services 0.9% 1.0% 0.07% 4.4% 4.4% 0.01% Other Industries 0.9% 1.0% 0.08% 5.1% 5.1% 0.01% Total 0.9% 1.0% 0.05% 4.2% 4.2% 0.01% 18

26 2.8 RESULTS The modelling here suggests that Australians forego substantial economic benefits as long as Indigenous disadvantage continues. The results in Table 2-9 show the cumulative percent changes (the sum of the annual changes) in the macroeconomic indicators by 2029 in the what if scenario compared with the base case. In aggregate, the modelling suggests improvements in Indigenous life expectancy, labour force participation rates and labour productivity will result in significant improvements to GDP and output per head. TABLE 2-9: RESULTS OF AE-GEM ANALYSIS 4 Real GDP will be around 1% higher than otherwise in In the absence of forecasts of GDP to 2029, 1% of GDP in is around $10 billion Shock Macroeconomic Indicator % Change Real Gross Domestic Product Value-Added Productivity Capital Stock Employment Consumer Price Index Real Household Consumption Population Labour Supply Indirect Tax Base Real Government Consumption Real Investment Real Exports Real Imports Terms of Trade Rate of Return on Capital Government Price Index Real Wage Further, since the percentage change in GDP is greater than the percentage change in the population (which is 0.2% higher in 2029), living standards (measured in terms of output per head) also rise. Real wages per person will be lower in 2029, by around 0.2%. This outcome arises as there is a large increase in the total labour supply, but a smaller increase in total labour productivity. This smaller increase means productivity per worker falls and, therefore, real wages also fall. Although wages per person fall, total wages paid to workers will increase. Realisation of the economic benefits is predicated on a higher proportion of the Indigenous population in paid work, able to take on higher skilled and better paid work and working 4 Most of the indicators improve except the terms of trade, the consumer price index, and the government price index. Since Indigenous employment and earnings are higher than otherwise, real household consumption is also higher (0.86%). While imports are consequently higher (0.69%), the price of imports is unchanged (ie. the world price). On the other hand, the supply of exports increases (0.77%) and the price of Australian exports falls, reducing the terms of trade (0.26%). The rate of expansion in the labour force, together with improvements in labour productivity outweigh increases in demand, so the Consumer Price Index (CPI) is 0.26% lower than otherwise. The government price index is lower than otherwise for similar reasons to the CPI, except that labour costs are a higher proportion of the government sector than other industries. 19

27 full-time rather than part-time. Improvements in the health and educational attainment of Indigenous people are fundamental to achieving these results STATIC VERSUS DYNAMIC We have undertaken a simplified analysis comparing two points in time based on linear growth patterns in all variables. In particular, we have not attempted to explain how the end point is achieved. As discussed in the next chapter, policies, programs, skills and resources will be required to assist Indigenous Australians to overcome the disadvantages they face. Box 3-2, Box 3-3, and Box 3-4 illustrate the types of analyses and actions required at a micro level and the complex nature of the likely interventions needed REGIONAL EFFECTS Regional (geographic) effects which are relatively important for analyses of Indigenous populations were not in scope for this project. It is possible that under the scenario modelled, the expansion in the Indigenous labour supply and improvement in labour productivity may ameliorate labour shortages in regional areas, and contribute to economies of scale in some smaller regional cities INCOME DISTRIBUTION While distributional effects are not specifically modelled, an increase in the labour supply and employment rate of Indigenous Australians enhances their ability to share in economic prosperity. However, the implicit modelling assumption is that income distribution improves because Indigenous educational attainment and employment rates improve. Chart 2-8 details the current quintiles of mean equivalised gross household income per week distributed across levels of educational attainment, for both the Indigenous and the total Australian population (for the year ). At every grouping of highest educational attainment level, a higher proportion of the total Australian population achieves the highest income quintile than the Indigenous population. For example, in the highest household income quintile, the equivalised mean 5 is $1,239 (as at , ABS 2007a). At the nonschool qualification level (i.e. tertiary education, vocational employment training, apprenticeships etc.), 30% of the total population achieves the highest income quintile, compared to only 10% of the Indigenous population. A reverse trend is observed in the lower income quintiles (i.e. higher comparative Indigenous weightings). Chart 2-9 analyses the change in the current income quintile breakdown by educational attainment level should the indigenous population reach an equivalent breakdown by 2026 to that of the Australian average. In order for the Indigenous population to reach the Australia wide trends, the Indigenous per annum mean equivalised household income growth rate will need to increase by 1.16 percentage points (from 2.42% p.a. base case to 3.58% p.a. what if scenario). At an aggregated level, the change to the overall national average weekly income growth rate is 0.02 percentage points (from 2.39% p.a. base case to 2.41% p.a. what if scenario). 5 An equivalised mean corresponds to a single individual s mean gross income (i.e. household income is divided by the number of people over 15 years of age in the household). 20

28 CHART 2-8: COMPARATIVE POPULATIONS INCOME QUINTILE* DISTRIBUTION BY HIGHEST EDUCATIONAL ATTAINMENT LEVEL 100% 80% 60% 40% 20% 0% % Pop (15+) Indigenous (36%) Australia (13%) Indigenous (24%) Australia (18%) Indigenous (9%) Australia (16%) Indigenous (32%) Australia (53%) Year 9 or below Year 10 or Year 11 Year 12 A non-school qualification Lowest ($255) 2nd ($414) 3rd ($565) 4th ($746) Highest ($1,239) Source: SCRGSP 2007 and ABS 2007a. *Estimated by mean equivalised gross household income per week, values per income quintile displayed in legend. CHART 2-9: AVERAGE WEEKLY INCOME BASED ON HIGHEST LEVEL OF EDUCATIONAL ATTAINMENT AND EQUIVALENT INCOME PROFILES $1,400 $1,200 $1,000 $800 $600 $400 $200 Australian (Base) Australian (shock) Indigenous (Base) Indigenous (shock) $ Forecast Year (as at 30 June) Source: Calculations by access economics based on SCRGSP 2007 and ABS 2007a. 21

29 3. A CHANGE IN THE STRUCTURE OF GOVERNMENT BUDGETS In this chapter, the impact on government budgets of matching the life expectancy and other characteristics of Indigenous people with the national average is discussed. The aim is to project revenue and expenditure in 2029 based on different assumptions about Indigenous life expectancy, labour force participation, labour productivity and expenditure per head for key programs in a number of portfolios. Revenue is discussed first, followed by expenditure. The methodology is discussed in detail in Appendix A. All expenditure and revenue estimates are expressed in 2009 dollars. 3.1 INTRODUCTION Improvements in life expectancy, labour force participation and productivity of Indigenous people will change the structure of government budgets. For example, a priori, it would be expected that: as the participation rates of Indigenous children in school (particularly in the latter years of secondary school) increase to match the Australian average, Australian, State and Territory Government expenditure on school education will increase; extending the life span of Indigenous people will contribute to increased expenditure on ageing related programs, such as the Aged pension and residential aged care; and conversely, if the health of Indigenous Australians reaches a level commensurate with that of all Australians, government expenditure overall on health will fall, largely because government expenditure on hospitals will fall although Medicare and pharmaceutical outlays are likely to rise. In a similar but more restricted exercise, Taylor and Hunter (1998) estimated that if the Indigenous unemployment rate was reduced to that of the rest of the population, the savings to government in payments to the unemployed, in 1996 dollars, would be around $193 million by the year 2001 and $274 million by 2006 with unemployment bills of $112 and $126 million respectively. Tax receipts from achieving parity in labour force status would approximate $177 million by Their analysis was based on 1991 and 1996 Census data, with the 1991 data adjusted for changed propensity to identify as Indigenous. Importantly 1 While improvements in Indigenous life expectancy, labour force participation, and labour productivity over the next 20 years may enhance the budget bottom line in some key mainstream programs, such improvements are only likely to be achieved with additional government outlays on Indigenous-specific or other types of assistance programs. These types of outlays have sometimes been called remedial outlays in the literature. Some examples are discussed in Box 3-2, Box 3-3Box 3-4, and Box 3-3Box 3-4. Some of the key Indigenous-specific programs are noted here, but generally not included as part of the structural changes to government budgets. 2 All of the analysis in this chapter assumes other influences on government revenue and expenditure remain constant. This enables analysis of the impact of improving the circumstances of Indigenous people alone. However, over a 20 year period, other factors affecting government budgets are also likely to change. 22

30 3 Regional (geographic) analysis is out of scope here, however, for various reasons, the structure and efficiency of programs may differ according the geographic distribution of the recipients. Other factors may also influence the nature of spending within some mainstream programs. For example, an increase in the need to accommodate cultural and other characteristics of Indigenous Australians in program delivery over time if Indigenous people are to access mainstream programs at a greater rate. 4 Taxes and government outlays (transfer payments) do not use resources, but simply redistribute the available income. Hence, savings to government budgets do not reflect changes in total population wealth. However, transfer payments are associated with economic costs (called deadweight losses) such as the costs involved in administration of welfare payments systems, and the distortion of individual work and leisure choices attributable to government programs. Deadweight losses are not incorporated in the analysis here. 3.2 REVENUE Forecast changes to several important areas of government revenue are in Table 3-1. All other factors are held constant in other words, the increased revenue is wholly attributable to reducing Indigenous disadvantage and does not reflect other influences that could also potentially affect government income. All numbers are in 2009 dollars. In 2029, with an improved outlook for Indigenous Australians, government revenue overall (including Australian, State and Territory) is projected to be approximately $4.6 billion higher than otherwise INCOME TAX Income tax is the Australian Government s largest single source of income (with revenue from income tax expected to be $105.5 billion in 2009 Table 3-1) and the largest source of revenue gain ($1.7 billion higher in 2029). After adjusting for the Medicare levy, Family Tax Benefit and Child Care Benefit, the net gain to revenue from personal taxes is around $1.9 billion. This reflects the higher population, higher employment to population ratio, and shifting across progressive tax brackets in 2029 in the what if scenario compared with the base case GOODS AND SERVICES TAX (GST) The most significant source of revenue for State and Territory Governments is the goods and services tax (GST). In the modelled scenario, in 2029 GST revenue is projected to be $530 million higher than otherwise, reflecting higher household consumption. The AE-GEM forecasts suggested household consumption would be 0.8% higher in 2029 than otherwise in the what if scenario and GST revenue is projected to increase by a similar proportion. 23

31 TABLE 3-1: IMPACT ON GOVERNMENT REVENUE ($2009 MILLION) Government Income Revenue % of Total Base case Income in (2009 $ million) after changes to 2009 Govt Revenue 2029 Indigenous LE and LFP Amount of revenue rise State Government Personal Gambling $5, % $6,840 $6,910 $70 1.0% Subtotal $5, % $6,840 $6,910 $70 1.0% Business Payroll tax $15, % $19,840 $20,000 $ % Subtotal $15, % $19,840 $20,000 $ % Consumption and Other Land and Property $16, % $21,930 $22,120 $ % GST $49, % $63,820 $64,350 $ % Financial and Capital Transactions $14, % $18,720 $18,830 $ % Provision of Goods and Services $ % $350 $350 $0 0.0% Insurance $2, % $3,810 $3,840 $30 0.8% Motor Vehicle Taxes $7, % $9,420 $9,490 $70 0.7% Other $ % $680 $680 $0 0.0% Subtotal $91, % $118,720 $119,670 $ % State Total Income $112, % $145,400 $146,580 $1, % Australian Government Personal Income tax $105, % $124,790 $126,480 $1, % Other Personal $30, % $35,960 $36,310 $ % Medicare Levy $8, % $9,750 $9,830 $80 0.8% Refunds $21, % -$25,880 -$26,090 -$ % Subtotal $166, % $144,620 $146,530 $1, % Business Corporate tax $67, % $79,800 $80,580 $ % Superannuation $8, % $9,740 $9,820 $80 0.8% Fringe Benefits tax $4, % $4,940 $4,980 $40 0.8% Subtotal $79, % $94,480 $95,380 $ % Consumption Sales/Other Indirect tax $1, % $1,480 $1,490 $10 0.7% Excise $24, % $30,940 $31,240 $ % Subtotal $25, % $32,420 $32,730 $ % Other Petroleum Resources Rent Tax $1, % $1,580 $1,590 $10 0.6% Customs Duties $6, % $8,770 $8,850 $80 0.9% Other taxes, fees and fines $2, % $3,530 $3,560 $30 0.8% Non-tax revenue $15, % $20,390 $20,580 $ % Subtotal $26, % $34,270 $34,580 $ % Total Australian Government $298, % $305,790 $309,220 $3, % Total Income $410, % $451,190 $455,800 $4, % Source: modelling by Access Economics Rise as % of base case

32 3.3 EXPENDITURE Box 3-1 Expenditure assumptions As explained above in the introduction to this chapter, while remedial outlays are likely to be necessary if improvements in the circumstances of Indigenous Australians are to be achieved, they not included in this analysis. The estimates here are based on the set of mainstream programs currently in place. Some Indigenous specific programs are, however, noted. The methodology is described in detail in Appendix A. In brief, two scenarios are estimated for the year 2029 similar to those in the previous chapter a base case and a what if scenario. In the base case, 2009 and 2029 government expenditure per head (and total expenditure) is estimated based on base case Indigenous population estimates and the continuation of current Indigenous participation rates (for example in education), welfare recipient rates or government expenditure per head. In the what if scenario, 2009 and 2029 government expenditure per head (and total expenditure) is generated based on the what if scenario Indigenous population estimates, assuming Indigenous participation rates (for example, in education), welfare recipient rates, or government expenditure per head gradually match the Australian average WELFARE PAYMENTS (AUSTRALIAN GOVERNMENT) In , 52% of the Indigenous population reported government pensions as their primary source of income, close to twice the rate of the total population, which reports government pensions as their primary source of income at 26% (SCRGSP 2007, p.186, attachment table 3A.6.8). The following welfare payments funded and administered by the Australian Government are discussed here: Newstart, Youth Allowance (job seekers), Parenting Payment, Sickness Allowance, Disability Support Pension, Aged Pension, Carer s Payment, Carer s Allowance, and CDEP. Total expenditure on these programs is detailed in Chart

33 CHART 3-1: TOTAL EXPENDITURE (ALL AUSTRALIANS) INCORPORATED IN THIS ANALYSIS, $ MILLION, 2007 $25,000 $20,000 $15,000 $10,000 $5,000 $0 Newstart Allowance Youth Parenting Allowance Payment Partnered Parenting Payment Single Sickness Allowance Disability Support Aged Pension Carer's Payment Carer's Allowance CDEP (2008) If Indigenous life expectancy, labour force participation and labour productivity in 2029 are at the same levels as today (i.e. the base case), government outlays on this set of programs would be $2.4 billion (Table 3-2). By comparison, if the outlook for Indigenous people in 2029 is much the same as the Australian average, (i.e. the what if scenario) government outlays in 2029 would be almost half this $1.4 billion with a saving of $1.1 billion. Savings are projected for all programs considered here except the Aged Pension (Table 3-2). In the what if scenario, outlays on Aged Pension payments in 2029 are higher than in the base case because of improvements in Indigenous life expectancy. In 2029, there are nearly 15,000 more Indigenous people aged 65 years and over alive in the what if scenario than in the base case. The largest source of savings in dollar terms in 2029 derives from the Newstart Allowance (including Newstart, Mature Age Newstart and Partner Allowance) a saving of $473 million in that year. This is because in 2009, 8.6% of Newstart recipients are Indigenous over three times their representation in the population. People aged over 65 years are not eligible for Newstart, so increased Indigenous life expectancy has a relatively minor effect, with the eligible population forecast to be only 4.7% higher in 2029 under the what if scenario than the base result, a difference of 19,900. In contrast, the projected number of Indigenous recipients of Newstart is 21,306 lower in the what if scenario than the base case reflecting lower Indigenous unemployment due in part to improved education and health. Other forms of welfare which are likely to generate savings are also analysed. Sickness, Disability and Carers payments which when the gap is closed are likely to be less than in the base case because the number of individuals who are able to work outweighs the overall increase in the population. The Youth Allowance which is likely to be less than otherwise because of lower youth unemployment among Indigenous people which outweighs the impact of higher school retention rates of Indigenous students who claim the Youth Allowance while completing their studies. 26

34 Savings in Single and Partnered Parenting Payments are generated because individuals previously eligible in 2029 are no longer eligible because their incomes are higher and they fail the means tests. The estimates of savings from CDEP should merely be noted. At the time of writing, governments were still considering the future of this program, and the need for an employment training initiative will depend on a number of factors including job opportunities in remote areas. Savings arise under the methodology applied here because CDEP is an Indigenous-specific program, with 99% of recipients identifying as Indigenous. 27

35 TABLE 3-2: WELFARE PROJECTIONS AND SAVINGS, (2009 DOLLARS $ 000) (A) Program Base Shock* Savings Newstart Allowance Indigenous population base 235, , , , , ,550 Indigenous expenditure ($'000) $344,640 $348,390 $373,300 $403,820 $629,220 $156,450 $472,770 Average per capita spend $460 $430 $410 $370 Indigenous per capita spend $1,460 $1,440 $1,510 $1,550 $1,550 $370 Youth Allowance Indigenous population base 84,230 85,030 90,050 98, , ,560 Indigenous expenditure ($'000) $76,280 $91,050 $86,040 $93,100 $110,750 $20,840 $89,910 Average per capita spend $900 $200 $200 $170 Indigenous per capita spend $900 $1,070 $960 $940 $940 $170 Parenting Payment Partnered Indigenous population base 265, , , , , ,690 Indigenous expenditure ($'000) $71,030 $74,730 $78,640 $84,410 $126,720 $39,100 $87,620 Average per capita spend $100 $90 $90 $80 Indigenous per capita spend $270 $280 $280 $290 $290 $80 Parenting Payment Single Indigenous population base 43,090 45, , , , ,690 Indigenous expenditure ($'000) $269,180 $292,780 $303,500 $330,350 $495,910 $167,650 $328,260 Average per capita spend $380 $370 $350 $360 Indigenous per capita spend $1,010 $1,090 $1,090 $1,120 $1,120 $360 Sickness Allowance Indigenous population base 43,090 45, , , , ,550 Indigenous expenditure ($'000) $1,850 $2,000 $1,990 $2,170 $3,380 $2,670 -$2,670 Average per capita spend $7 $7 $7 $6 Indigenous per capita spend $8 $8 $8 $8 $8 $6 Disability Support Indigenous population base 43,090 45, , , , ,410 Indigenous expenditure ($'000) $175,280 $283,120 $318,810 $375,840 $550,310 $320,590 $230 Average per capita spend $410 $610 $620 $650 Indigenous per capita spend $610 $970 $1,050 $1,170 $1,170 $650 Aged Pension Indigenous population base 43,090 45,480 34,800 20,030 60,000 66,730 Indigenous expenditure ($'000) $84,970 $92,390 $109,670 $128,650 $333,980 $562,720 -$228,740 Average per capita spend $7,580 $7,660 $8,250 $8,430 Indigenous per capita spend $5,230 $5,400 $6,060 $6,420 $6,420 $8,430 Carer's Payment Indigenous population base 43,090 45, , , , ,690 Indigenous expenditure ($'000) $33,580 $39,970 $49,670 $64,550 $97,720 $56,100 $41,620 Average per capita spend $70 $70 $80 $100 Indigenous per capita spend $110 $120 $150 $180 $180 $100 Carer's Allowance Indigenous population base 43,090 45, , , , ,690 Indigenous expenditure ($'000) $35,070 $41,200 $47,590 $59,820 $90,560 $51,920 $38,640 Average per capita spend $70 $80 $80 $90 Indigenous per capita spend $110 $130 $140 $170 $170 $90 TOTAL EXPENDITURE ($'000) $1,091,880 $1,265,630 $1,369,210 $1,542,710 $2,438,550 $1,378,040 $1,057,130 CDEP Indigenous population base 505, , , , , ,060 Indigenous expenditure ($'000) $117,350 $121,170 $174,860 $4,720 $170,140 % Indigenous population on CDEP 3.6% % national population on CDEP 0.1% 0.1% 3.6% *The shock column represents the what if scenario the movement due to increased life expectancy, and labour force participation and productivity changes. (a) Method described in detail in Appendix A. 28

36 Box 3-2 Child abuse and neglect In , Indigenous children were more than five times as likely as other children to be the subject of a substantiation 6 of abuse or neglect (AIHW 2008a). Other reports have highlighted some of the complex underlying causes for the over-representation of Indigenous children in the child protection system (eg. NTBIPACSA 2007, HREOC 1997, Young 2006). The impact of child abuse and neglect on children and their families and communities is considerable including: pain and suffering; premature death arising either from the injury, or from consequent long term mental illness and potential for suicidal behaviour; health system costs of injury or mental illness; the costs to governments of administering child protection and out of home care services; learning difficulties and associated educational support; lost productivity victims of abuse are more likely to be unemployed or underemployed; crime (including juvenile delinquency, adult criminality, homelessness, substance abuse); and intergenerational transmission of abuse. Keatsdale (2003) estimated the cost of all of these impacts for the year at $4.93 billion (including both Indigenous and non-indigenous Australians). 7 The authors estimated there were around 38,700 abused and neglected children in so the cost was $127,386 per abused child. If Indigenous children are more than five times as likely to be the subject of abuse or neglect and Indigenous people comprise 2.4% of the population, a ball park estimate of the cost of abuse and neglect of Indigenous children is $592 million in Inflated at 2.5% per year, this is around $686 million in The costs of ameliorating child abuse and neglect A detailed analysis of the types of actions necessary to ameliorate child abuse is out-of-scope here, but notably: the Northern Territory Government response to the Little Children are Sacred Report (NTBIPACSA 2007) was costed at $286.4 million over five years 8 ; and the Australian Government s Northern Territory Emergency Response will involve expenditure in excess of $580 million in Substantiation means that a child protection agency has investigated a report of child abuse or neglect, and has concluded that the child has been, is being, or is likely to be, abused, neglected or otherwise harmed. 7 On Keatsdale s (2003) estimates, abuse prevention programs represented only around 3% of the total cost. 8 NTCMD FaCSIA

37 3.3.2 HEALTH EXPENDITURE (AUSTRALIAN, STATE AND TERRITORY GOVERNMENTS) Table 3-3 provides a breakdown of expenditure per capita by broad health service and Indigenous status. In , spending per head on Indigenous Australians was substantially higher than the Australian average for: public hospitals; community health services (reflecting expenditure on Indigenous specific health services Aboriginal Community Controlled Health Organisations); patient transport; and public health research and other health services (including aids and appliances) (AIHW 2008). Conversely, spending per head on Indigenous Australians was substantially less than the Australian average for: Medicare; pharmaceuticals; residential aged care; dental services; and private hospitals. TABLE 3-3: HEALTH EXPENDITURE PER CAPITA, BY BROAD HEALTH SERVICE AND INDIGENOUS STATUS, AND 2009 State and Federal Government Expenditure (2005 $'000) Expenditure per person (2005 $) Expenditure per person (2009 $) Australian Australian Broad Health Service Total % Indigenous Indigenous Average Ratio Indigenous Average Ratio Private hospitals $2,549, % $30 $ $40 $ Public hospitals 1 $20,625, % $2,050 $1, $2,600 $1, Medical sevices $11,588, % $280 $ $360 $ Community health services 2 $3,530, % $990 $ $1,250 $ Dental and other professions $1,598, % $70 $ $90 $ Pharmaceuticals $6,051, % $140 $ $180 $ Services for older people $4,392, % $60 $ $80 $ Patient transport $1,398, % $200 $ $260 $ Public health research $2,779, % $240 $ $310 $ Other health services 3 $2,097, % $150 $ $190 $ Total health expenditure $56,611, % $4,210 $2, $5,360 $3, Includes emergency departments, other non-admitted patient services and public psychiatric hospitals 2 Includes funding of dental services by states and territories 3 Includes health administration, aids and appliances, and other health services Source: AIHW (2008) Overall, improving the life expectancy of Indigenous Australians and assuming per capita health expenditure on Indigenous people in 2029 is the same as that for all Australians leads to a saving of $1.3 billion in health expenditure in 2029 (Table 3-4). While outlays on Medicare, pharmaceuticals, residential aged care and private hospitals are higher than otherwise, these dissavings are outweighed by expenditure reductions in community health services and public hospitals. As Indigenous Australians earn higher incomes, they will be more readily able to afford private health insurance, and hence private hospital care. Higher Medicare expenditure may reflect greater use of Medicare funding by Indigenous Health Services, particularly in rural/remote areas. 30

38 TABLE 3-4: IMPACT ON GOVERNMENT HEALTH EXPENDITURE IN 2029(A) State and Federal Government 2029 Base 2029 Shock 2029 Broad Health Service Base ($'000s) Expenditure per Capita Shock ($'000s) Expenditure per Capita Savings ($'000s) Private hospitals $34,790 $40 $132,360 $160 -$97,570 Public hospitals 1 $2,077,460 $2,600 $1,070,620 $1,270 $1,006,840 Medical sevices $282,560 $350 $601,540 $710 -$318,980 Community health services 2 $1,000,320 $1,250 $183,270 $220 $817,050 Dental and other professions $71,190 $90 $82,950 $100 -$11,760 Pharmaceuticals $145,400 $180 $314,100 $370 -$168,700 Services for older people $60,330 $80 $228,000 $270 -$167,670 Patient transport $204,930 $260 $72,600 $90 $132,330 Public health research $246,760 $310 $144,290 $170 $102,470 Other health services 3 $154,050 $190 $108,860 $130 $45,190 Total health expenditure $4,277,790 $535 $2,938,590 $349 $1,339,200 (a) Method described in detail in Appendix A. 1 Includes emergency departments, other non-admitted patient services and public psychiatric hospitals 2 Includes funding of dental services by states and territories 3 Includes health administration, aids and appliances, and other health services Source: Calculations by Access Economics based on expenditure estimates in AIHW (2008). 31

39 Box 3-3 Estimate of remedial expenditure on primary care for Indigenous people Giving Indigenous Australians the same level of access to primary health care as non-indigenous Australians with comparable health status would require substantial increases in funding to expand the medical workforce. Access Economics (2004) estimated that at least 250 full time equivalent medical practitioners would be required as well as additional access to the PBS. Overall, it was estimated that funding for primary care would need to increase by around $400 million per annum. In addition, training more health professionals would be necessary at a cost of $36.5 million per annum (running to $167 million over six years). The workforce estimates were derived by calculating how many additional primary care (medical) services Indigenous people are likely to need based on their higher prevalence of long term health conditions and their greater use of hospitals. Although the Indigenous population is younger than the non- Indigenous population, the average health of Indigenous Australians was similar to that of other Australians aged 50 to 54 years. The rate at which non- Indigenous people aged 50 to 54 used services was therefore suggested as the benchmark for the number of Medicare services per head Indigenous people should be using. On this basis, Indigenous Australians need to use Medicare billed medical services per person each year. Further, to kick-start the disease prevention cycle among Indigenous people, it was suggested that there should be a mark-up on service use by Indigenous Australians. Such a strategy would involve the careful targeting of key risk factors with a view to reducing both morbidity and mortality. In the absence of empirical research, it was assumed that a gross addition of 20% to the primary care workforce would lead to a net requirement of 15%, as increased targeting of prevention is likely to lead to offsetting savings in tertiary care (Access Economics, 2004:9). Mooney (1998) made a similar argument suggesting that topping up funding for Indigenous health based on differences in current levels of need may do no more than stop the health gap widening. Other studies have similarly suggested a mark-up is necessary because of the need for Aboriginal and Torres Strait Islander services to provide culturally appropriate care, including the need for cross-cultural mediation. Thus, the final Access Economics (2004) estimate was that Indigenous people need 14.7 medical services per person year EDUCATION (AUSTRALIAN, STATE AND TERRITORY GOVERNMENTS) Indigenous Australians presently have much lower participation and attainment rates in education than the national average, and these gaps increase with the level of schooling. For example, in 2006 apparent retention rates for full-time students between years 10 and 12 were 76.1% for the total population, but only 46.7% for Indigenous students (SCRGSP 2007, attachment table 4A.117), while 2005 participation rates in preschool education are also lower for Indigenous children, with 23.8% of preschool aged Indigenous children enrolled compared to a national average of 28% (SCRGSP 2007). Table 3-5 shows that Indigenous students in 2005 lagged behind the national average rate of attainment of all educational benchmarks by 10%, and that this gap increases with the level of schooling. 32

40 TABLE 3-5: ATTAINMENT OF NATIONAL BENCHMARKS, 2005 Comparisons of base case and what if scenario projections of expenditure on education in 2029 are consistent with a priori expectations that total education spending would need to be higher than otherwise to achieve improvements in Indigenous peoples life expectancy and living standards (Table 3-6). Government outlays on education are $273 million higher in 2029 in the what if scenario. The largest source of dissaving in 2029 is Higher education ($265.3 million). In 2009, the Indigenous Higher education participation rate is 3.1%, 63% of the national average (4.9%). There is a forecast saving in Vocational Education and Training (VET) expenditure of $206.7 million in Currently, Indigenous people have higher participation rates in VET than the rest of the population which falls to match the national average in 2029 reflecting an increase in participation in other forms of education (secondary school and university). TABLE 3-6: IMPACT OF EDUCATION PARTICIPATION ON GOVERNMENT BUDGETS(A) (a) Method and sources described in detail in Appendix A. For noting Indigenous-specific education funding programs are not included in the analysis but may be required to achieve increased Indigenous participation in education. Some such Indigenousspecific education funding programs are already in place. Reading Writing Numeracy Year 3 Total 92.7 ± ± ± 1.1 Indigenous 78.0 ± ± ± 3.8 Year 5 Total 87.5 ± ± ± 1.3 Indigenous 62.8 ± ± ± 3.9 Year 7 Total 89.8 ± ± ± 0.9 Indigenous 63.8 ± ± ± 2.9 Program Expenditure ($'000s) Base Shock* Savings Expenditure on Indigenous Students by Level of School Preschool $100,900 $150,190 $252,990 -$102,800 Primary and Secondary $1,410,920 $1,710,410 $1,825,010 -$114,600 Higher education $196,490 $308,600 $527,480 -$218,880 VET $211,980 $332,930 $144,540 $188,390 Total Expenditure on Indigenous Students $1,920,290 $2,502,130 $2,750,020 -$247,890 The Indigenous Education Strategic Initiatives Program (IESIP) includes recurrent supplementary assistance for Indigenous students, remedial English as a second language work with students whose first language is their traditional Indigenous language, and projects under the National Indigenous English and Numeracy Strategy. This Program provides funding to schools based on remoteness and initiatives taken by the schools to promote positive Indigenous education outcomes. Funding for IESIP is undertaken on a quadrennial basis. In 2004 funding for IESIP totalled $41.2 million. The Indigenous Education Direct Assistance Program (IEDAP) includes the Aboriginal Tutorial Assistance Scheme, Aboriginal Student Support and a Parent Awareness Scheme. It further provides a Vocation and Educational Guidance Scheme. Funding 33

41 is based upon a per capita formula which takes into account the number of Indigenous students enrolled at the school, their level of schooling and a remoteness weighting. In 2005 funding for IEDAP totalled $67.7 million. In addition, the Australian Government also provides income support to students (Austudy including Youth Allowance payments, and Abstudy). It is difficult to forecast likely scenarios for these payments. Using the methodology outlined in detail in Appendix A, under the what if scenario, Government expenditure on income support for students would be less than otherwise, probably reflecting higher Indigenous incomes not meeting means test requirements for eligibility (Table 3-7). Most of the savings arise under the Abstudy program. However, this type of remedial expenditure is likely to be necessary to achieve increases in Indigenous participation in post secondary education, so projecting savings is not necessarily consistent with reality and reflects the uncertainty surrounding this type of analysis. TABLE 3-7: IMPACT OF EDUCATION RELATED WELFARE ON GOVERNMENT BUDGETS(A) Program Expenditure ($'000s) Base Shock* Savings Abstudy Total Expenditure $222,240 $214,290 $318,650 $61,190 $257,460 Average per capita expenditure $10 $10 Average indigenous per capita expenditure $670 $600 $600 $50 Austudy Total Expenditure - Austudy and Youth Allowance $3,870 $13,340 $20,410 $6,950 $13,460 Average per capita expenditure $110 $160 $50 $10 Average indigenous per capita expenditure $30 $50 Total Indigenous education expenditure 1 $226,110 $227,630 $339,060 $68,140 $270,920 (a) Method and sources described in detail in Appendix A JUSTICE (STATE AND TERRITORY GOVERNMENTS) Indigenous Australians are currently over-represented in Australia s prison population and therefore government expenditure on the prison system. In 2007 there were on average 15,723 Indigenous Australians serving some form of imprisonment or community correction sentence on any given day, compared with the population total of 79,416. Indigenous Australians therefore accounted for 19.8% of the total population serving criminal sentences. 2.98% of Indigenous Australians were serving some form of correction order on any given day of 2007, compared to just 0.38% of the total national population, meaning Indigenous Australians were over-represented in the Australian prison population at a rate of almost 8 to 1. It is anticipated that closing the gap of Indigenous disadvantage would produce a corresponding fall in the imprisonment rate of Indigenous Australians, and that there are therefore large potential savings from such a reduction in the prisoner population. Table 3-8 shows the results of estimates on government expenditure on justice in 2029 and the potential savings from the what if to Indigenous living standards. The largest potential savings occur in secure prisons. The total saving is $870.3 million in 2029, with spending on secure prisons 90.8% less than otherwise because there are fewer Indigenous prisoners. 34

42 TABLE 3-8: IMPACT OF JUSTICE ON GOVERNMENT BUDGETS (A) Correction Type Base Shock* Savings Open plus Periodic Detention Indigenous Expenditure ($'000) $123,990 $135,510 $205,150 $19,430 $185,720 Total Expenditure ($'000) $593,520 $593,460 Indigenous per capita spend $370 $380 $380 $30 Average per capita spend $40 $30 Secure Indigenous Expenditure ($'000) $386,070 $429,560 $650,320 $58,810 $591,510 Total Expenditure ($'000) $1,667,910 $1,796,470 Indigenous per capita spend $1,150 $1,210 $1,210 $100 Average per capita spend $100 $100 Community Corrections Indigenous Expenditure ($'000) $45,660 $50,950 $77,130 $8,390 $68,740 Total Expenditure ($'000) $248,040 $256,330 Indigenous per capita spend $140 $150 $150 $20 Average per capita spend $20 $20 Total Indigenous Expenditure ($'000) $555,720 $616,020 $932,600 $86,630 $845,970 (A) Method and sources described in detail in Appendix A. 35

43 Box 3-4 Petrol sniffing Petrol sniffing has devastating health and social consequences, causing (among other things) mental impairment, brain injury and death. While both Indigenous and non-indigenous people engage in petrol sniffing, it has been relatively more prevalent among Aboriginal and Torres Strait Islander communities. Ameliorating sniffing Most importantly, ameliorating sniffing among Indigenous communities would extend life: the average age at death of sniffers is around 23 years for Indigenous people, this implies an average of 42 years of life lost per addict. In addition, some of the costs to governments and communities of sniffing that would be avoided if sniffing were ameliorated include: expenditure on hospitalisations from sniffing related illness (at an average cost of $4,855 per episode of hospital care in 2005), on rehabilitation, and on long term care and accommodation ($59,532 per person in 2005) of sniffers with permanent or long term brain damage; government outlays on unemployment benefits. Addicts are less likely to be employed than other Indigenous people. In one Australian Indigenous community, eradication of petrol sniffing increased the likelihood of sniffers being employed from 7% to 63%. Temporary absences from work due to sniffing related illness would also be reduced; justice system expenditure including police and court time, incarcerations and inquests; and the costs of caring for addicts incurred by community, family and friends. In the proposed roll out region for Opal fuel in Central Australia in 2005, the crime and justice system costs of petrol sniffing were estimated at $16.2 million, productivity losses at $8.3 million, health, long term care and rehabilitation costs were each about $4.1 million, $4.2 million and $3.7 million respectively, and informal care provided by families and significant others $2.3 million. The costs of ameliorating sniffing The economic costs of petrol sniffing would only be avoided through well targeted strategies to prevent the take up of sniffing and reduce the resultant adverse outcomes. Suggested strategies involve expanding government outlays and include: supply of appropriate fuel (possibly requiring government subsidies); and the implementation of successful education, harm minimisation, legal deterrence and employment and training programs, together with community action. The total costs of a package of such programs in the Opal fuel rollout region was estimated to cost $26.6 million in Source: Access Economics (2006) 36

44 3.3.5 HOUSING (AUSTRALIAN, STATE AND TERRITORY GOVERNMENTS) There are a number of housing programs provided by Australian, State and Territory Governments. The focus here is on mainstream public housing, State owned and managed Indigenous housing, mainstream community housing, Indigenous community housing and Commonwealth Rent Assistance. Indigenous Australians currently access all of these at much higher rates than the total Australian population. Consequently, a reduction in Indigenous disadvantage is likely to be associated with savings to governments in housing related programs. Public and community housing Overall, projected savings in 2029 are over $173 million (Table 3-9) most from public housing (savings of $170.4 million). Indigenous Australians currently access public housing at four times the rate of the overall population. Notably, the method in Table 3-9 assumes that the proportion of the population accessing Indigenous specific housing does not change, so expenditure on these programs in 2029 is higher than otherwise because of the higher Indigenous what if scenario population in Indigenous specific housing programs may be considered remedial expenditure. TABLE 3-9: PUBLIC AND COMMUNITY HOUSING RELATED CHANGES TO GOVERNMENT BUDGETS (A) Program Expenditure ($'000s) Base Shock* Savings Indigenous Community Housing % Indigenous Population 26% 26% 26% % Total Population 0% Total Indigenous Expenditure $62,970 $91,790 $96,880 -$5,090 National Community Housing % Indigenous Population 2.6% 2.6% 0.9% % Total Population 0.9% Total Indigenous Expenditure $16,530 $24,090 $8,800 $15,290 State Owned and Managed Indigenous Housing % Indigenous Population 17% 17% 17% % Total Population 0% Total Indigenous Expenditure $85,350 $124,420 $131,320 -$6,900 National Public (State Owned) Housing % Indigenous Population 33% 33% 8% % Total Population 8% Total Indigenous Expenditure $159,900 $233,090 $62,670 $170,420 Total Indigenous housing expenditure $324,750 $473,390 $299,670 $173,720 (A) Method and sources described in detail in Appendix A. Commonwealth Rent Assistance (CRA) Indigenous Australians currently access CRA at close to twice the rate of the total population. If Indigenous recipient rates match those of the Australian average, savings to the Australian Government are projected at around $40 million in 2029 (Table 3-10). 37

45 TABLE 3-10: RENTAL ASSISTANCE CHANGES TO GOVERNMENT BUDGETS (A) Commonwealth Rental Assistance Expenditure ($'000s) Base Shock* Savings % Indigenous Population 9.2% 8.8% 8.8% 5.4% % Total Population 5.6% 5.4% Total Indigenous Rental Assistance $69,510 $75,840 $114,820 $75,640 $39,180 (A) Method and sources described in detail in Appendix A. 38

46 4. BURDEN OF DISEASE The implications of improving the circumstances of Indigenous Australians for GDP, output per Australian, and for government budgets do not include other important reasons underlying arguments for action. A key companion to the economic improvements modelled in other chapters of this report is a reduction in the burden of disease experienced by Indigenous Australians, and thus an improvement in their health and quality of life. Differences in the burden of disease experienced by Indigenous Australians compared with the Australian population as a whole are discussed in this chapter based on the work of Vos et al (2007) and Begg et al (2007). 4.1 OVERVIEW AND DEFINITIONS Vos et al (2007) compared the burden of disease in the total Australian population in 2003 as calculated by Begg et al (2007) with that in the Indigenous population in More than 170 diseases and injuries are included in the comparison. Health related quality of life is described using burden of disease measures. Specifically, Vos et al (2007) and Begg et al (2007) used disability adjusted life years (DALYs) to measure the burden of disease. DALYs are described in Box 4-1. To find the excess burden of disease among Indigenous people, the Indigenous burden of disease is compared with the values the DALY estimates would have taken had Indigenous Australians experienced the same mortality and disability as the general Australian population. Since the disease burden varies with age, and the age distribution of the Indigenous population is different to that of the total Australian population (Indigenous people are on average younger), comparisons between the Indigenous and all Australian population are standardised for age. The estimated gap in life expectancy in this chapter differs from that in previous chapters of this report because a different methodology has been used. The life expectancy gap between the Indigenous and total Australian population estimated by Vos et al (2007) and used to estimate the difference in the disease burden between the Indigenous and total Australian population is less than that estimated by the ABS (see Box 4-1) the ABS estimated the gap at around 17 years compared with the Vos et al (2007) estimates of approximately 13 years. If the ABS estimates had been used, the burden of disease among Indigenous Australians compared with the total Australian population would have been greater than calculated by Vos et al (2007) and presented here. The modelling estimates in Chapters 2 and 3 above are based on the ABS life expectancy gap estimates. If the Vos et al (2007) estimates had been used, the impact on output over 20 years would have been lower. 39

47 Box 4-1 Measuring the Burden of disease DALYs Vos et al (2007) utilise Disability Adjusted Life Years (DALYs) as a measure of the disease burden. One DALY is equivalent to one year of healthy life lost, including both fatal and non-fatal disease: DALY = YLL + YLD Where YLL represents years of life lost due to premature death, and YLD represents a year lived with disability. YLL = number of deaths x expected life remaining (in years) YLD = number of incident cases x disability weight (range 0-1) x duration of disability (in years) Expected life remaining It is important to note that Vos et al (2007) estimated the life expectancy at birth of Indigenous Australians for the period 1996 to 2001 to be 64 years for males and 69 years for females, a gap of 12.5 and 13.5 years with life expectancy of the total Australian population, respectively. These life expectancy estimates are higher than those reported by the ABS for the same period. There is a scientific debate about the validity of either set of estimates that can only be resolved when new and better data and methods become available. If the ABS mortality figures had been adopted, the total Indigenous population burden of disease estimates would have been higher (since the ABS estimates a larger gap in life expectancy for the total Indigenous Australian population). Disability weights To calculate years of life lost due to disability, disability weights are applied as per the equation above. A disability weight of one represents death, and a weight of zero represents a year of healthy life. The weights are derived from a survey of clinicians and public health experts. For example, according to the weighting system, on average, society judges a year with blindness (weight 0.43) to be preferable to a year with paraplegia (weight 0.57), and a year with paraplegia to be preferable to a year with unremitting unipolar major depression (weight 0.76) (Begg et al (2007)). Source: Vos et al (2007) and Begg et al (2007). 4.2 BURDEN OF DISEASE The disease burden among Indigenous Australians occurred at a considerably higher rate at each age compared with the total Australian population. In 2003, the Indigenous Australian population made up 2.4% of the total Australian population; however, despite its much younger age structure, the Indigenous Australian population carried 3.6% of the total disease burden (Vos et al 2007). Among Indigenous people in 2003: 40

48 cardiovascular disease and mental disorders (including substance use disorders) were the leading causes of disease burden, together accounting for 32% of the disease burden; and chronic respiratory disease, diabetes mellitus and cancers were the next three leading causes, accounting for around 8% each of the disease burden. Comparisons of Indigenous health and total Australian population health can be made from the findings of Vos et al (2007). Cancer was responsible for a much smaller proportion of the disease burden among Indigenous people (19% total Australians compared with 8% among Indigenous people). Diabetes, and unintentional and intentional injuries were each responsible for a larger proportion of the total burden in Indigenous Australians than in the total Australian population. The proportion of the disease burden accounted for by premature death was higher among Indigenous Australians (54%) than for all Australians (49%) once ill, Indigenous people are more likely to die prematurely than other Australians. Figure 4-1 (reproduced from Vos et al (2007)) depicts the differences in the distribution of the burden of disease by age. The majority of the absolute burden (the number of DALYs on the left hand axis) for Indigenous Australians occurred in the middle-aged population with a significant peak also occurring in the very young. In the total Australian population, the absolute burden continued to increase into old age. DALYs per 1000 people (on the right hand axis) occurred at a considerably higher rate at each age for Indigenous Australians compared with the total Australian population. Table 4-1 (reproduced from Vos et al (2007)) shows the differences in the distribution of burden by condition between the Indigenous and total Australian populations. For Indigenous Australian males, otitis media, homicide and violence, birth trauma and asphyxia, and low birth weight caused substantially more of the disease burden than in the total Australian population. For Indigenous females, sexually transmitted diseases, homicide and violence, otitis media, and rheumatic heart disease were more important contributors to the burden of disease than for all Australian females. 41

49 FIGURE 4-1 DALYS BY AGE AND SEX, INDIGENOUS AUSTRALIAN AND TOTAL AUSTRALIAN POPULATIONS, 2003 Source: Vos et al, (2007) 42